|Home | About | Journals | Submit | Contact Us | Français|
Using rat organotypic hippocampal-entorhinal cortical (HEC) slice cultures, we examined whether phospholipase A2 (PLA2) activity is involved in binge alcohol (ethanol)-induced neurodegeneration, and whether docosahexaenoic acid (DHA; 22:6n-3), a fish oil-enriched fatty acid that is anti-inflammatory in brain damage models, is neuroprotective. Assessed with propidium iodide and lactate dehydrogenase (LDH) leakage, neurodamage from ethanol (6 days 100 mM ethanol with four withdrawal periods) was prevented by the PLA2 pan-inhibitor, mepacrine. Also, ethanol-dependent neurodegeneration— particularly in the entorhinal region—was significantly ameliorated by DHA supplementation (25 µM); however, adrenic acid, a 22:4n-6 analog, was ineffective. Consistent with PLA2 activation, [3H] liberation was approximately fivefold greater in [3H]arachidonic acid-preloaded HEC slice cultures during ethanol withdrawal compared to controls, and DHA supplementation suppressed [3H] release to control levels. DHA might antagonize PLA2 activity directly or suppress upstream activators (e.g., oxidative stress); however, other DHA mechanisms could be important in subdueing ethanol-induced PLA2-dependent and independent neuroinflammatory processes.
Chronic binging alcoholics incur brain neurodegeneration , but the underlying mechanisms are unresolved. While there are in vitro indications for excitotoxicity [2, 3], in vivo attempts to prove that augmented N-methyl-d-aspartate (NMDA) receptor activation underlies ethanol-induced neurodegeneration have been generally unsuccessful. Binge ethanol-intoxicated adult rats exhibit selective neurodegeneration in the olfactory (entorhinal) cortex, notably layer III pyramidal neurons, and the hippocampus, particularly dentate granule cells [4–6]. Neurodamage in these regions is not significantly diminished by blocking NMDA receptors, voltage-gated calcium channels, or nitric oxide synthase activity [7–9]. However, modest but significant brain edema and electrolyte imbalances occur during the repetitive ethanol intoxication, and diuretics—e.g., furosemide  or acetazolamide —in suppressing the edema, exert neuroprotection.
One pathway linked to edema/swelling during trauma or ischemia in brain (esp. astroglia) is phospholipase A2 (PLA2) activation, which leads to disproportionate liberation/mobilization of arachidonic acid (AA, 20:4 n-6) . Elevated AA generates reactive oxygen species (ROS), pro-inflammatory eicosanoids, and other metabolites that promote neuroinflammation and neurodegeneration ; then again, in a positive feedback-like scenario, oxidative stress can be upstream of (further) PLA2 activation [13, 14]. Also, AA itself has its direct (pro)apoptotic effects [15–17]. Parenthetically, it is relevant to mention that certain ROS trapping agents protect against binge ethanol-induced neurodegeneration in vivo [8, 18].
In contrast to n-6 AA, long-chain n-3 polyunsaturated fatty acids such as docosahexaenoic acid (DHA, 22:6), enriched in fish oils, are frequently associated with neuroprotection and neuroinflammatory suppression . Of significance, brain concentrations of DHA are reduced by chronic ethanol [20–22]. We examined whether PLA2 inhibition with mepacrine (MEP) or DHA supplementation in our in vitro model of binge ethanol-induced brain damage could reduce neuronal death, assessed principally by propidium iodide (PI) labeling of degenerating neurons , and, if so, whether there was an association with AA liberation at early stages of ethanol exposure and withdrawal.
These experiments examining the effects of a PLA2 inhibitor and of DHA utilized long-term cultures of rat organotypic slices of hippocampus and entorhinal cortex, regions mentioned above that are susceptible to binge ethanol-induced neurodamage in rats . Organotypic brain slice cultures have advantages over mixed brain cultures in retaining localized neuronal-glial relationships . Furthermore, unlike slices of hippocampus that are often used in brain slice cultures, hippocampal-entorhinal cortical (HEC) slices in long-term culture have functional perforant pathways , which may be important for mechanisms underlying neuronal vulnerability and protection. Our findings with binge ethanol-treated organotypic HEC cultures indicate a role for PLA2 activation and, as in other brain damage models, DHA exerts significant neuroprotection that could involve containment of AA release and associated PLA2-related neuroinflammation, as well as other mechanisms to be investigated.
Unless otherwise indicated, supplies were obtained from Fisher (Pittsburgh, PA) and chemicals from Sigma (St. Louis, MO). Rat organotypic HEC slices were prepared from Sprague–Dawley 1-week-old pups (Harlan, Indianapolis, IN) according to Collins et al. , as modified from Stoppini et al. . Pups were sacrificed with minimal suffering according to a Loyola Medical Center IACUC-approved protocol. With tissue encompassing the entorhinal/hippocampal complex, transverse slices (350 µm) were taken and placed on 0.4 µm Millicel tissue culture inserts (Millipore, Bedford, MA; 3–4 slices/insert) in 6-well Falcon plates with covers. They were cultured in an atmosphere of 95% O2/5% CO2 at 37°C in MEM media (1 ml/well) containing 25% horse serum (Gibco Invitrogen, Carlsbad, CA) and 6.5 mg/ml glucose; media was changed every 3–4 days. Slices were examined before ethanol or control treatment (below) and those appearing unhealthy as indicated by significant PI labeling or obvious gliosis were discarded.
After 16–18 days in culture, slices were treated over a 6-day period with fresh media with or without 100 mM ethanol, a level not unusual in chronic alcoholics [26, 27], according to a “binge” schedule: ethanol-containing media for three consecutive 24 h days, followed by the first overnight withdrawal (ethanol-free media), ethanol 12–14 h daily for the following 2 days, each followed by an overnight withdrawal, and a fourth ethanol exposure for a day and overnight, followed by a ~ 2 h withdrawal, for a total of four withdrawal periods. Although the plates were in Tupperware containers, ethanol concentrations were permitted to diminish during exposure to similate the in vivo situation. With selected cultures MEP (1 µM), DHA (25 µM; Cayman, Ann Arbor, MI) or adrenic acid (ADA, 25 µM; Cayman, Ann Arbor, MI)—an n-6 polyunsaturated fatty acid of the same carbon chain length—was added ~ 3 h prior to the start of ethanol exposure (pre-incubation) and was supplemented throughout ethanol + withdrawal. PI, extensively used to assess specific neurodegeneration in organotypic slices in culture [23, 28], was added to culture media (5 µg/ml) during the last hr of the fourth withdrawal. Measurement of intensity of neuronal PI uptake was accomplished in HEC slices using Image J analysis software and a Nikon DS-5 camera + Nikon TS100 inverted microscope. The percentage of entorhinal cortical (EC) area with PI fluorescence above background was calculated in relation to the total area for each slice; the same threshold level was used for all slices in a given experiment. To verify PI results in the ethanol/MEP experiments, LDH activity—a neurotoxicity measure in slice cultures that correlates with PI labeling —was assayed in media pooled from the final day of ethanol + withdrawal using a Sigma LDC kit and standardized to slice protein, assayed with a bicinchoninic acid method (Pierce, Rockford, IL).
In a modification of an AA release procedure in HEC slices , [3H]AA (#0196, ARC, St. Louis, MO; 200 Ci/mmol) was added to experimental and control HEC slice culture media (1 µCi/ml media) 6 h before initiation of ethanol treatment. Measurement of selected HEC tissue extracts from rinsed slices obtained immediately prior to initial ethanol addition indicated that tissue [3H] content, which constituted ~ 70% of the added [3H]AA in agreement with other culture studies , did not significantly differ between fatty acid-supplemented and control slices. In media collected 20 min into the first withdrawal, [3H] content representing mainly released AA and AA oxidation products  was assessed by scintillation spectroscopy and normalized to protein, determined as above in homogenates of HEC slices. Mean values for PI labeling results and media [3H] were compared with one-way ANOVA and an appropriate post-hoc test as indicated in the figure legends.
Figure 1 shows representative images of experimental HEC slices in culture, carried through the experimental procedure as described and examined for PI labeling of degenerating neurons. Control HEC slices (Cont) carried through the 6-day protocol including media changes but without ethanol showed minimal PI-labeling. Figure 1B is a control slice exposed to MEP throughout the procedure that incurred no apparent increase over controls in neuronal PI labeling. Likewise, slices supplemented throughout with DHA (top rt) showed no noticeable increases in PI-labeled cells relative to control slices. In Fig. 1D, binge ethanol treatment (Fig. 1E, 100 mM) of HEC slices over 6 days with withdrawals evoked extensive neurodegeneration as indicated by PI labeling of neurons in the EC and the hippocampal dentate gyrus (DG) regions. However, HEC slices exposed to the ethanol + withdrawal protocol and supplemented as described with either PLA2 inhibitor MEP (MEP + E; image in Fig. 1E with slice rotated ~ 90°) or n-3 DHA (Fig. 1F; DHA + E) demonstrated evident reductions in PI-labeled degenerating neurons in both the EC and DG regions of the slices.
In Fig. 2 is shown quantitation of neurotoxicity in HEC slice experiments with 100 mM binge ethanol in the absence and presence of MEP. In Fig. 2A, quantitation of PI-labeled EC neurons represented in Fig. 1 demonstrated that binge ethanol treatment caused extensive neurodegeneration. MEP alone caused no increase in neurodegeneration, but its addition effectively prevented most of the binge ethanol-induced neurodamage (MEP + E). Further confirmation of cytotoxicity due to binge ethanol and significant reduction by MEP co-treatment is in Fig. 2B, which shows the effects of these treatments on LDH leakage in HEC slice culture. In accord with earlier HEC slice culture LDH results , binge ethanol treatment caused significantly increased media LDH activity (E) relative to controls (Cont), and this leakage representing degenerating cells was significantly—but not completely—suppressed by MEP co-treatment (MEP + E).
Figure 3 displays the effects on binge ethanol-induced neurodegeneration (PI labeling of EC neurons) of supplementation with two polyunsaturated fatty acids, DHA and ADA, a 22:4 analog that at times has been used as a n-6 polyunsaturated fatty acid comparison for DHA. In Fig. 3, pre-incubation/supplementation with DHA (25 µM), while having no effects on control slice PI labeling, completely blocked (E + DHA) the EC neurodegeneration caused by binge ethanol treatment (E) of HEC slices. Comparatively, Fig. 3 shows that supplementation with 25 µM n-6 ADA had no significant effect (E + ADA) on EC neurodegeneration in the HEC slice cultures. DG cell PI labeling in the ethanol-treated slices, while clearly suppressed by MEP or DHA (Fig. 1), was not quantitated.
The box plots in Fig. 4 demonstrate the effect of the neurotoxic binge ethanol exposure on tritium release from [3H]AA-pre-incorporated HEC slices during the first of the withdrawal periods, and the effect on release of neuroprotective DHA pre-incubation. Assayed in media taken 20 min into the first withdrawal period, the [3H] results showed, when compared to release from control (Cont) cultures, robust (approximately fivefold) release of [3H] associated with ethanol withdrawal (E) that indicates significant activation of PLA2 activity due to binge ethanol. Subsequent withdrawal periods also showed increased [3H] release compared to controls, but to lesser extents (not shown). Pre-incubation/supplementation with DHA as in Fig. 3 did not alter basal (control) [3H] release, but it completely abrogated the increase in [3H] release due to binge ethanol (E + DHA), signifying that DHA supplementation effectively suppresses binge ethanol-dependent activation of PLA2-dependent mechanisms.
The findings show that the neurodegeneration provoked in organotypic HEC slices by subchronic binge ethanol exposure involves augmented PLA2 activity as evidenced by the extensive neuroprotection from a general PLA2 inhibitor (MEP), and the [3H] release experiments further suggest that substantially elevated mobilization of n-6 AA, a well-documented neuroinflammatory accomplice, occurs early in the binge ethanol protocol. This is the first experimental data to our knowledge that directly implicates PLA2 activity with “binge alcoholic brain damage,” and current studies with selective inhibitors are underway to determine the specific PLA2 forms involved . We also find that supplementation with n-3 DHA—but not n-6 ADA, a 22-carbon elongation product of AA—affords essentially complete neuroprotection in concert with blockade of the induced AA mobilization. These results are consistent with binge or episodic ethanol-induced brain damage involving to an appreciable extent neuroinflammatory PLA2 activation, excess AA mobilization and oxidative stress that are conceivably downstream of neuroglial edema/electrolyte dysregulation [7, 34]. Brain (esp. cellular) swelling is known to increase PLA2 activity [11, 35, 36], and in positive feedback-like fashion, excessively released AA can potentially instigate more brain edema [37, 38] as well as increase oxidative stress (ROS)—which can trigger further PLA2 activation .
On the other hand, when supplemented or potentiated, n-3 polyunsaturated fatty acids—in particular, DHA—frequently have neuroprotective, anti-inflammatory, and survival effects [19, 39]. Much of the anti-inflammatory evidence for DHA is in vivo; however, the molecule has been linked to neuroprotection in various brain and other culture models as well [40–43]. Pertinent to our case is a study with rat organotypic hippocampal slice cultures in which DHA pre-incubation protected against ischemia-induced brain damage . Also, the n-3 fatty acid had positive effects on survival, neurite outgrowth and neuronal differentiation in brain cultures [45, 46], and suppressed epileptiform activity in acute hippocampal slices . As mentioned, ADA, the 22-carbon n-6 fatty acid elongation product of AA that is enriched in human brain plasmalogens , was not effective in suppressing ethanol-induced neurodamage. Unlike the potent neuroprotective effect of DHA, which is incorporated to a great extent into neuronal membrane , ADA’s lack of neuroprotection might in part relate to its high incorporation in myelin [50, 51] and consequently its relatively slow turnover. We acknowledge that a more appropriate (although not essential at this point) comparison for DHA is docosapentaenoic acid (22:5 n-6), a fatty acid shown to be reciprocally related to DHA in neuronal membranes  and which we plan to use in future studies.
Several possibilities might explain DHA’s inhibition of induced liberation of AA in DHA-supplemented brain slices during neurotoxic binge ethanol exposure. There are indications, albeit limited, that supplemented DHA can inhibit PLA2 ; and conversely, reduced tissue DHA levels (dietary n-3 deprivation) have been associated with elevations in rat brain cortical secretory and cytosolic PLA2 isoforms . Since supplemented DHA is incorporated and released by PLA2 isoforms along with AA, as demonstrated in pathological situations [13, 55], it is also possible, that suppressed AA release in DHA-supplemented slices during binge ethanol treatment reflects competition with incorporated DHA; however, since AA may be released by a different PLA2 isoform than DHA, as reported with ATP-stimulated glia , such competition might not be relevant. Parenthetically, it is possible that inhibition by MEP actually increases endogenous neuroprotective DHA. A further mechanism is that through its potential antioxidative effects , DHA could counter oxidative stress that in some studies is upstream (in addition to downstream) of PLA2 activation [13, 14]. Potentially related to this possibility is that DHA neuroprotection from glutamate toxicity in brain cultures or rat brain in vivo is associated with increased glutathione, antioxidant enzymes, and stabilization of intracellular calcium [40, 58], anti-oxidative actions that could affect PLA2 activity in our DHA-supplemented ethanol model. An additional AA-related mechanism for DHA is its reported inhibition of cyclooxygenase-mediated AA metabolism , which would lessen formation of neurodeleterious eicosanoids, as well as derived ROS.
However, mechanisms for DHA neuroprotection against ethanol-induced neurodamage other than or in addition to abrogation of excessive AA liberation and/or metabolism could well be important and, if so, it would suggest alternative neurotoxic mechanisms for binge ethanol as well. DHA incorporation specifically into neuronal membrane phosphatidylserine (PS) has recently been demonstrated , and research linking this phenomenon to neuronal survival reviewed and summarized . In brief, prolonged ethanol exposure can deplete neuronal PS  as well as DHA within membrane PS . Through its ability to increase PS in inner leaflets of neuronal membranes, added DHA can facilitate translocation of PKB/Akt, a key anti-inflammatory kinase which antagonizes apoptotic signaling . An equally intriguing possibility is that free or liberated DHA can be oxygenated (apparently in glia) by lipoxygenase-like enzymes to neuroprotectin-1 and resolvin molecules, which have marked—possibly receptor-mediated—neuroprotective properties [63–65]. These possible neuroprotective mechanisms for DHA need to be evaluated in future studies of the n-3 fatty acid’s propitious actions on binge ethanol-induced neurodegeneration. Regardless of these latter possibilities, the significantly increased [3H] mobilization from incorporated [3H]AA during binge ethanol treatment (further analyses will ascertain the AA content of the released [3H]) and its suppression by neuroprotective doses of DHA remain consistent with a model in which PLA2, AA and potentially derived ROS have roles in binge ethanol-dependent neurodegeneration.
The PLA2-based mechanistic scheme in Fig. 5 depicts how diuretic treatment, inhibition by MEP, DHA acting possibly at several steps, and ROS trapping agents might exert neuroprotection. Also added is the aforementioned possibility of upstream activation (dashed arrow) of PLA2 isoforms by ROS , enabling a “neurotoxic cycle.” For completeness, Fig. 5 includes as well the possibility that ROS might arise during binge ethanol exposure via increased membrane NADPH oxidase activity, and/or from brain mitochondrial damage and ROS leakage. The first of these latter two possibilities is indicated, for example, by evidence that preconditioning-like neuroprotection against ischemia/reperfusion that is exerted by an acute ethanol dose in gerbils involves NADPH oxidase-derived ROS . Additionally, mitochondrial dysfunction and derived ROS due to toxic ethanol treatment, mostly in liver but also in brain cultures, have been reported [67, 68]. However, whether these two latter pathways are linked to ethanol-induced brain edema and cell swelling is an open question.
High ethanol intake or exposure is linked to reductions in brain DHA levels [20, 22]. There are several possible reasons, and an important one is that ethanol can potentiate DHA oxidation , contributing to its brain depletion. Added to this are epidemiological studies showing that binge-drinking individuals tend to have lower n-3 dietary intakes than nondrinkers . Given DHA’s essential role in brain cognition [71, 72], its potential deficit in alcohol abuse could be a factor underlying increased vulnerability of the brain to synaptic and neuronal dysfunction. In association with others , we concur that dietary DHA or fish oil supplementation might afford a significant element of neuroprotection as well as functional cognitive benefits in active alcoholics.
Supported by the Loyola University Medical Center (LUMC) Alcohol Research Program (NIH T32 AA13527) and an LUMC Potts award to M.A.C.